System and Method for Predicting Industrial Equipment Motor Behavior

1. A non-transitory computer-readable medium with instructions stored thereon, wherein the medium is configured to predict a behavior of a motor, that when executed by a processor, performs the steps comprising: determining a plurality of predictive machine learning models; receiving input values from at least one sensor, said at least one sensor is configured to receive operational data from the motor; updating a probability of a plurality of operational outcome of the motor for each of a unit time stamp within a predetermined time period, comprising: calculating a percentage of input values that are correct; removing, at random, a plurality of input values corresponding to the percentage that are incorrect; calculating a mean square prediction error value from the input values not removed; calculating a weight of each of the plurality of predictive machine learning models; determining an upper bound and a lower bound of the input values not removed; calculating a counter value for each of the plurality of predictive machine learning models; updating the weight of each of the plurality of predictive machine learning models; calculating a normal probability; updating the upper bound and lower bound of the input values not removed; updating the normal probability based on the updated upper bound and updated lower bound of the input values not removed; calculating the probability of a plurality of operational outcomes of the motor; predicting the behavior of the motor for said unit time stamp; and transferring data respecting the predicted behavior of the motor to at least one of a personal computer (PC) and a smart phone.

2. The non-transitory computer-readable medium of claim 1, wherein calculating the weight comprises calculation of the weighted average contribution of each of the plurality of predictive machine learning models to a total of the plurality of predictive machine learning model.

3. The non-transitory computer-readable medium of claim 2, wherein updating the weight of each of the plurality of predictive machine learning models comprises updating said weight in every time stamp period.

4. The non-transitory computer-readable medium of claim 1, wherein the method is used to predict thermal runaway.

5. The non-transitory computer-readable medium of claim 1, wherein the method is used to generate a range of temperature range predictions.

6. The non-transitory computer-readable medium of claim 1, wherein the processor generates an urgent alert signal if thermal runaway is predicted.

7. A system comprising: a non-transitory computer-readable medium with instructions encoded thereon, wherein the medium is configured to predict a behavior of a motor; and one or more processors configured to, when executing the instructions, perform operations of: determining a plurality of predictive machine learning models; receiving input values from at least one sensor, said at least one sensor is configured to receive operational data from the motor; updating a probability of a plurality of operational outcome of the motor for each of a unit time stamp within a predetermined time period, comprising: calculating a percentage of input values that are correct; removing, at random, a plurality of input values corresponding to the percentage that are incorrect; calculating a mean square prediction error value from the input values not removed; calculating a weight of each of the plurality of predictive machine learning models; determining an upper bound and a lower bound of the input values not removed; calculating a counter value for each of the plurality of predictive machine learning models; updating the weight of each of the plurality of predictive machine learning models; calculating a normal probability; updating the upper bound and lower bound of the input values not removed; updating the normal probability based on the updated upper bound and updated lower bound of the input values not removed; calculating the probability of a plurality of operational outcomes of the motor; predicting the behavior of the motor for said unit time stamp; and transferring data respecting the predicted behavior of the motor to at least one of a personal computer (PC) and a smart phone.

8. The system of claim 7, wherein calculating the weight comprises calculation of the weighted average contribution of each of the plurality of predictive machine learning models to a total of the plurality of predictive machine learning model.

9. The system of claim 7, wherein updating the weight of each of the plurality of predictive machine learning models comprises updating said weight in every time stamp period.

10. The system of claim 7, wherein the method is used to predict thermal runaway.

11. The system of claim 7, wherein the method is used to generate a range of temperature range predictions.

12. The system of claim 7, wherein the processor generates an urgent alert signal if thermal runaway is predicted.